基于视频的高速公路车辆检测和跟踪算法

Vehicle detection and tracking algorithm based video for freeway

采用一种基于像素的新方法来更新背景,同时将工作区域分为车辆检测和车辆跟踪两个部分,在车辆检测区建立跟踪对象,在车辆跟踪区仅仅对跟踪对象进行预测跟踪,从而减少了跟踪的盲目性和算法的计算量,提高算法精度。根据该算法设计的视频检测器已用于实际交通参数的采集,统计的结果表明该算法检测和跟踪的正确率分别高于96%和97%,并具有很好的实时性,能满足管理部门的需求。

To reduce the blindness in tracking and improve the speed of performance and the correct rate, a new algorithm for vehicle detection and tracking was presented. In the algorithm, the background was updated with a new method based pixel and the detection region is divided into two parts： vehicle detection region and vehicle tracking region. The objects tracked were constructed in the vehicle detection region and only tracked in the vehicle tracking region through their previous states. The video detectors based on the algorithm have been used to collect traffic parameters, and experimental results show that the correct rates of vehicle detection and vehicle tracking are respectively large than 96% and 97%.